Classics and Archaeology Virtual Museum

Report on Progress, 13 December 1998

The Virtual Museum Goes On-line!

The Virtual Museum Project website, with a draft version of the Classics and Archaeology Virtual Museum, is now on-line. The current (temporary) URL of the website is The home page contains a link to the database, and from there you can access the images.

A simpler, easier-to-remember address is being developed. There is a link from the home page of the School of Fine Arts, Classical Studies and Archaeology. This link will be kept up to date when the Virtual Museum address changes.

Please note that the on-line version of the database and image archive is functional but still a limited draft only. It is searchable, but its functionality will be further extended. Coins and manuscripts have yet to be integrated into the main object database. Only about a dozen fields of information about each object have been put on-line (the full version has over 100 fields), and even those fields need further information or editing. Most of the images are quite small and still in a fairly raw state. But there is at least one image of every object (except for missing objects), and all of the image links are active.

During testing it was discovered that the web server software eliminated paragraph breaks when serving the contents of a database field on the web. This made the longer field entries very difficult to read. Further investigation and experimentation resulted in a solution to this problem.

In addition to the on-line Virtual Museum database and image archive, the draft website contains:

back to top

Photography Completed

The last of the photography has now been completed - including the final 200 coins and the box of bronzes which turned up after we had almost despaired of ever finding them. We can only hope that the story of the other missing objects will have an equally happy ending. Also included in the very last shoot was a handsome piece of Roman glass from the Vizard Foundation antiquities collection, also professionally restored and conservated by the Ian Potter Art Conservation Centre.

The six Photo-CDs resulting from the last round of photography have been indexed and processed. The objects were re-packed and carefully carried across from the studio to the storeroom in Old Arts. Shelving and recording of their new locations will be completed when an opportunity presents itself.

Many thanks to the Multimedia Education Unit, which contributed about two days per week since photography commenced in December 1996. And particularly to the team of photographers and image processors, who combined top professional skills with efficiency, creativity, patience, open-minded interest and a wonderful spirit of co-operativeness.

back to top

On-line Images Number 6,879

4,856 raw images have been processed in the last few weeks, bringing the total number of processed images to 6,879. All of these images are now accessible from the web-mounted database.

Processing involves copying each source image from the Photo-CD to a hard disk, opening the image in Photoshop, reducing the size of the image so that it will download quickly over the internet, and saving the image in the compressed jpeg format which web browsers can display. Each image is then given a unique identifying name of the type xxxxyyyz.jpg, where the first four characters (x) correspond to the last four digits of the Photo-CD 12-digit unique identifying code, the next three characters (y) represent the number of the image on the Photo-CD, the eighth character (z) is an in-house letter code for the size of the image, and the suffix (.jpg) tells computers about the format of the file and the form of compression used. This new naming system has been devised to ensure that the image file names are totally platform-independent, eliminating some problems experienced earlier when copying files from a Macintosh to a Unix machine and from there to a DOS-based or Windows system.

This processing has been done in order get at least a small version of each image accessible via the web-mounted database. These versions will be replaced progressively over the next few months by edited images. Editing will produce images that are much larger (just a little smaller than a standard computer screen) but still small enough in file size to download quickly on the web. It involves cropping out 'dead' space so that the object fills the display window and, where necessary, rotating the object so that it displays the right way up. It can also involve making adjustments to contrast, hue and saturation levels etc. As each image has to be edited individually, it is an extremely labour-intensive task.

back to top

Web-database Interface Re-designed

The web-database interface has been extensively re-designed. Users can now search for either a single object or a set of objects.

In both cases, you can search for all objects that match ANY of the search criteria, or only for objects that match ALL of the search criteria you select. When searching for a single object, however, you can actually find a set of objects because the search engine has been set to allow truncated criteria and to find all objects containing tthe characters you type. Entering '1987' in the 'accession no.' box will thus produce a list of all objects accessioned in 1987 and every 1987th object accessioned in any year.

The search for a single object assumes that you know at least part of one of the various identifying numbers associated with an object:

Searching for a set of objects does not require you to know or enter any information. Instead, the database generates a drop-down list of all entries in the relevant field and you select the entry you want. For the draft version of the Virtual Museum the following search fields have been included:

You can thus find all the bronzes, or only the bronzes from Palealona. The results of the search can be sorted according to any available field. The objects found are listed with their accession number, type, and the period of the work. Objects without an accession number (on loan, missing) have been given a dummy accession number '0000'. The number of results displayed has been restricted to 40 per screen, in order to keep the download time reasonable (feedback would be welcome on this point).

Clicking on the accession number brings up a new browser window with details of that particular object. For the time being, the fields supplied include only the following sixteen: primary category, secondary category, type, ware, dimensions, findspot, material, collection, bibliography, credit line, accession no, other no., previous no., site registration no., period of work and date of work. This information in these fields is not yet complete for every object; and additional fields will be added as they are edited. Also included in the details is a list of file names of images of the object, with an indication of the size of the image file; clicking on the file name brings up the image in an independent window. To facilitate comparisons, more than one image window can be open at the same time.

Feedback about the design and usefulness of the web-database interface would be most welcome. 

back to top

Object Database Revised for On-line Appearance

All the fields currently displayed from the web-based object database have been edited and given a more consistent format; some new fields have been added to the database.

The most important aspect of this task was to decide upon a standard classification system from the many that have been implemented at various times throughout the history of the separate collections that comprise the Virtual Museum. Diverse types of objects, such as a papyrus, a trefoil oinochoe and a strip of gold leaf, do not naturally fit into a single set of descriptive categories. The solution adopted has been to accept that all fields are not necessarily relevant to all objects, and that for most objects there will be some fields that are redundant.

Even a casual user will soon notice the many gaps in the information that is currently displayed, and the inconsistencies that remain (for example, a fully rationalised dating system has yet to be finalised). These gaps and anomalies will be eliminated over the next few months as collaborators supply the relevent information or edit and approve the information that has already been gathered but not yet confirmed.

The version of the object database that has been mounted on the web is a cut-down version. Fields and layouts not currently displaying have been stripped out, and indexing has been kept to the minimum needed in order to keep the database as small as possible.

It is proposed to update the displayed object database on a monthly basis.

For the use of those collaborating with the Project a fuller version of the database will be maintained on-line at a different, private address. This version will be kept up to date on a daily basis after it has been brought into line with the changes necessitated by mounting the public version on the world wide web.

back to top

Image Database Updated and Revised

In the last few weeks a further 1,748 records have been added to the image database. These represent the images on 18 Photo-CDs, each of which contains up to 100 images.

The source images on a Photo-CD are numbered 1-100. The record of each source image in the image database must therefore include the identifying number of the CD itself plus the sequence number of the particular image. A list of source images was created in a Microsoft Excel spreadsheet.

Indexing a Photo-CD involves opening up each image and identifying the object in the image. Though quite time consuming, in most cases this is straight forward because the accession number or another identifier has generally been included in the photograph. On some occasions, however, the wrong number got into the photograph, or the object was photographed with what has subsequently become its 'previous' number. The worst case involved a Cypriot bowl, one of about twenty in the collection, which not only looked like several others but had had its accession number revised without the old accession number being replaced on the object itself.

Once the object in each source image has been identified, its 'key' number in the object database has to be found and entered into the image database. Because there are objects that do not have an accession number, each object in the database has been assigned a unique 'key' number. This number is the essential link that relates a number of images to a single object. Having identified the object, therefore, the next step was to look up the object's key number in the object database and enter this into the spreadsheet. The contents of the spreadsheet were then imported into the image database, creating the required new records.

The source images were then copied to a hard disk and processed, as outlined above. Each processed image file was saved with a unique name and this name then entered into the image database along with the name of the folder in which the derivative image was stored. The image database can thus produce not just a list of the images of a particular object but also the location of those images on the web server (the filepath). The filepath is essential for the web browser to find the image.

Several types of check have been devised to ensure that each source image is correctly identified and entered accurately into the image database, but it is almost inevitable that more errors will be detected as people start working with the images and the object database.

A further step was to note the size, in kilobytes, of all 6,897 processed images and record this information in the image database. Despite some clever tricks using Hypercard and Excel, this too was a fairly time-consuming exercise. But it is regarded as best practice to display the size of files so that viewers on the internet will know how long it might take to download a file.

Ideally, each derivative image would be accompanied by a caption. It is hoped that future development of the project will allow for captions to be added to the database and displayed along with the image file names and file sizes.

back to top

Object Movies Number 101

A total of 101 object movies have now been completed and delivered. The raw files, taking up about 80 gigabytes of storage, have been archived on DAT tapes. The processed files occupy about 1.2 gigabytes and have been archived on two CD-ROM's, which have now been indexed. The active versions have been copied to hard disk.

As the movies occupy an average of 12 mb each, it is unlikely that there will be a high demand for them over the internet - it was always envisaged that they would be accessed primarily from the kiosk version of the Virtual Museum. Downloading them from on-campus is feasible but can still take more than a minute or two. Seven samples have been made available via the Virtual Museum website, along with instructions on how to obtain the software needed to view them via a web browser. The smallest file is that of the bronze cast of Poseidon, given to the University on the occasion of the Olympic Games in 1956.

Once storage issues on the web server have been sorted out, they will all be integrated into the image database.

back to top

Kiosk/development Machine Acquired

The final configuration of the kiosk machine was specified through an extensive consultation process with representatives of the Multimedia Education Unit, ArtsIT, Information Technology Services and the Ian Potter Museum of Art. This machine will be used for the further development of the project - especially the image editing - until the contents of the Virtual Museum are complete. It will then be set up as a kiosk and turned over to the Steering Committee which will determine its eventual location.

The core of the kiosk machine is a Power Macintosh G3 Minitower running at 333 megahertz. The RAM has been enhanced to 256 megabytes. Three internal hard disk drives bring total storage capacity to 27 gigabytes - the alternative was to have only two internal disk drives, but 18 gigabytes was felt to be just too close for comfort to the anticipated final size of the project and to leave no room for other materials. The hard drives will be formatted using Apple's new HFS+ architecture, which reduces storage overheads dramatically. The monitor is a 17" Apple multisync, which will allow images to be displayed at a much higher resolution than is possible over the web.

The hardware has been delivered, but setting up was delated until Version 4 of Norton Utilities was received (on 13 December). This is the only version that can repair HFS+ formatted hard disks.

Setting up also awaited release of the DAT tape drive from its full-time function as the archiving machine for the object movie raw files. The DAT drive will be attached to G3 to hold backups of the image files as they are edited. Backing up will take place at the end of each editing session, and automated using Retrospect.

Photoshop 5 has been purchased to speed up image processing and automate as much of it as possible. It is hoped that the new features of this version of Photoshop will permit image editing to attain standards higher than those defined in the project specifications.

It is hoped that the development machine can be set up, familiarization with Photoshop 5 completed, backup processes established and final image editing under way before Christmas.

back to top

Greek Vases Catalogue

Melbourne University Press eventually decided that the Catalogue of Greek Vases did not suit its list and declined to make an offer for its publication. On the other hand, Desktop Publishing and Design presented a quite reasonable quotation for printing the volume. Discussions with publishers continue and a decision is expected in the near future.

In the meantime, the recent visit of Professor Iakovides enabled consultation on the Mycenean vases with one of the leading experts on that period, and his advice was found to be very helpful.

back to top

Bab edh-Dhra'

The collections include a complete tomb group of pottery from Bab edh-Dhra', Jordan, which is believed to be the biblical Sodom. Information from the published excavation reports was deciphered, entered into the object database, and made available to Dr Bunnens for editing and supplementation.

The term 'deciphered' is apt. The excavation reports adopted a complex encoding system for classifying and describing the pottery from the site so that information about the finds could more easily be manipulated by computer. These codes were transcribed and a little computer routine devised to de-code them. Unfortunately, the key to a couple of the codes was not included in the excavation reports, and further information is needed before the decipherment can be completed.

Excavations in the area continue, and the excavators have set up a website with maps, photographs and an account of their progress. The Virtual Museum website now has a page devoted to Bab edh-Dhra', with an account of our pottery and links to the excavation website. There is also a link to a website devoted to the "Good Moose of Slime" - the author of which claims to have worked on the excavations at Bab edh-Dhra' and to have discovered nearby a cache of ancient texts, but some have suggested that it is no more than a parody of the Dead Sea Scrolls, which come from the near neighbourhood.

It is planned to supplement the databases with similar web pages containing background and current information about the various sites and objects in our collections. A method has been devised of including in the database active links to such pages from the detailed information served up about individual objects. It is hoped to implement this idea in the next monthly update of the on-line version of the Virtual Musuem.

back to top


Object Movies from the Virtual Museum were presented at the Doing IT at Melbourne conference on 5 November, the annual showcase for information technology and multimedia projects on campus. Congratulations to David Adam for a thorough presentation.

back to top

Hellenic Antiquites Museum

The Virtual Museum Project was represented at the opening of the Hellenic Antiquities Museum on 9 November by the Premier Mr Jeff Kennet and the Greek Minister for Culture Professor Venizelos. The opportunity was taken to visit the opening of the initial exhibition devoted to finds from the Mycenean tombs at Aidonia.

The possibility of an exhibition in 1999 of the University's Hellenic antiquities, and of the Virtual Museum, was mentioned in conversation with Dr Anna Malgorzevicz, Director of the Museum, and Professor D Penington.

An invitation was accepted to attend the lecture at the Museum by Professor S Iakovides on 25 November. At an earlier private visit to the University Professor Iakovides, a member of the Academy of Athens, provided advice in relation to the Univesity's Mycenaean holdings. He also indicated that, on his return to Greece, he would investigate the possibility of obtaining for the University some Mycenaean pottery sherds for teaching purposes.

back to top


The entry of information from the stocktake has continued, albeit mainly as a relief from processing images, developing web-database interfaces and the like. A further 224 records were updated. For convenience, the stocktake information has been entered into a separate mini database which will be integrated into the main database once all the information has been entered and checked.

back to top

Queen's College Collection

Dr Sagona has alerted the Project to the existence of the Queen's College collection, comprising about 40 antiquities from Egypt that were briefly listed by Dr Colin Hope about 15 years ago and kept in storage ever since. A proposal will be developed to include these objects in the Virtual Museum.

top of page